Abstract
Introduction
While patients with non-alcoholic fatty liver disease (NAFLD) are mostly overweight or obese, some are lean.
Methods
In a community-based follow-up study (baseline and follow-up surveys performed in 2007 and 2014), we investigated and compared the clinical characteristics, body composition, metabolic associations and outcomes, and other risk factors among individuals with lean (BMI < 23 kg/m2) NAFLD, non-lean (BMI ≥ 23 kg/m2) NAFLD and those without NAFLD. To investigate associations of selected genetic variants, we performed a case–control study between lean NAFLD cases and lean non-NAFLD controls.
Results
Of the 2985 participants in 2007, 120 (4.0%) had lean NAFLD and 816 (27.3%) had non-lean NAFLD. 1206 (40.4%) had no evidence of NAFLD (non-NAFLD). Compared to non-lean NAFLD, lean NAFLD was commoner among males (p < 0.001), and had a lower prevalence of hypertension (p < 0.001) and central obesity (WC < 90 cm for males, < 80 cm for females) (p < 0.001) without prominent differences in the prevalence of other metabolic comorbidities at baseline survey. Of 2142 individuals deemed as either NAFLD or non-NAFLD in 2007, 704 NAFLD individuals [84 lean NAFLD, 620 non-lean NAFLD] and 834 individuals with non-NAFLD in 2007 presented for follow-up in 2014. There was no difference in the occurrence of incident metabolic comorbidities between lean NAFLD and non-lean NAFLD. Of 294 individuals who were non-NAFLD in 2007 and lean in both 2007 and 2014, 84 (28.6%) had developed lean NAFLD, giving an annual incidence of 4.1%. Logistic regression identified the presence of diabetes at baseline, increase in weight from baseline to follow-up and a higher educational level as independent risk factors for the development of incident lean NAFLD. NAFLD association of PNPLA3 rs738409 was more pronounced among lean individuals (one-tailed p < 0.05) compared to the whole cohort sample.
Conclusion
Although lean NAFLD constitutes a small proportion of NAFLD, the risk of developing incident metabolic comorbidities is similar to that of non-lean NAFLD. A PNPLA3 variant showed association with lean NAFLD in the studied population. Therefore, lean NAFLD also warrants careful evaluation and follow-up.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
Introduction
Non-alcoholic fatty liver disease (NAFLD) is defined as hepatic steatosis, detected either on imaging or histology, in the absence of secondary causes [1]. NAFLD is probably the commonest chronic liver disease (CLD) worldwide [2]. The reported prevalence from Asia–Pacific countries ranges from 5 to 40% depending on the population studied and the method used to detect fatty liver [3].
Most patients with NAFLD are overweight (BMI ≥ 23 kg/m2) or obese (BMI ≥ 25 kg/m2) (non-lean NAFLD) [4]. However, some patients with NAFLD are lean (BMI < 23 kg/m2), and there has been an increasing clinical interest in the group [5,6,7,8]. A considerable proportion of Asians with NAFLD is described as having lean NAFLD. It was reported that present in ~ 20% of the Asian population, lean NAFLD is closely linked with insulin resistance and diabetes [8]. A recent multinational study showed an increase in mortality in lean individuals with NASH [9]. Another described its clinical characteristics and association with patatin-like phospholipase domain containing 3 (PNPLA3) risk allele carriage among Caucasians [10]. Despite these, lean NAFLD has failed to secure an unambiguous status as a unique disease entity [11].
The Ragama Health Study (RHS) is a large community-based cohort study on non-communicable diseases [12]. It is a collaborative study between the National Center for Global Health and Medicine (NCGM), Tokyo, Japan and the Faculty of Medicine, University of Kelaniya, Ragama, Sri Lanka. As part of this study, using stringent ultrasound criteria, we previously reported a community prevalence of 32.6% and an annual incidence of 6.6% for NAFLD in an urban, adult Sri Lankan population [12, 13]. We also reported a significant association between PNPLA3 gene rs738409 polymorphism and susceptibility to NAFLD in this population [14]. With such high rates in the community, cryptogenic (probably NAFLD related) CLD is very common in Sri Lanka [15].
Features of lean NAFLD among South Asians are poorly documented in the literature. Therefore, we investigated risk factors, including genetic polymorphisms, clinical characteristics, metabolic associations and outcomes, for lean NAFLD, and compared them with non-lean NAFLD and those without NAFLD in the RHS cohort after a 7-year follow-up period.
Methods
The RHS is a community-based, prospective, cohort follow-up study. The study population was chosen by age-stratified random sampling from electoral lists, from the Ragama Medical Officer of Health area in the Gampaha district on Sri Lanka. Initial screening was done in 2007 and follow-up was after 7 years, in 2014. The population consisted adults aged 35–64 years in 2007 who were 42–71 years in 2014. On both occasions, participants were screened by structured interview to collect socio-demographic variables and lifestyle habits (with special emphasis on patterns of alcohol consumption), collection of anthropometric indices, liver ultrasonography, and biochemical and serological tests.
A 10-mL sample of venous blood was obtained from each subject. This was used to determine, glycosylated hemoglobinA1c (HbA1c), fasting serum triglycerides (TG) and high density lipoproteins (HDL), serum alanine aminotransferase activity (ALT) and hepatitis B and C serology [hepatitis B surface antigen (HBsAg) and anti-hepatitis C virus antibodies (anti-HCV) using CTK Biotech ELISA kits]. Genetic analysis was done at initial screening only [15]. Body composition [Total body fat (TBF) and visceral fat percentage (VFP)] was assessed by a body composition monitor using a proven bioelectrical impedance method (Omron HBF-362 body composition monitor, Omron Healthcare, Lake Forrest, Illinois, United States) only at follow-up survey in 2014. Abnormal TBF definition for females was > 32% and for males > 25% while abnormal VFP for both females and males was defined as > 10% [16].
All subjects underwent ultrasonography of the liver with a 5-MHz 50-mm convex probe (MindrayDP-10 Ultrasound Diagnostic Systems, Mindray Medical International Limited, Shenzhen, China). Ultrasonographic examination was carried out by five doctors with special training in liver ultrasonography. NAFLD was diagnosed on standard USS criteria (two of the three following criteria: increased hepatic echogenicity compared to the spleen or the kidney, blurring of liver vasculature and deep attenuation of the ultrasonographic signal) and exclusion of alcohol overuse and other secondary causes such as hepatitis B and C. Normal cut-off values were based on the revised Adult Treatment Panel III (ATP III) criteria for metabolic syndrome for Asians [17].
At baseline and follow-up survey, lean NAFLD was defined by BMI < 23 kg/m2 and non-lean NAFLD was defined by BMI ≥ 23 kg/m2. It is widely accepted that the prevalent international criteria to define overweight (BMI > 25 kg/m2) and obesity (BMI > 30 kg/m2) are not suitable for Asian Indian and South Asian populations. Therefore, we adopted the suggested BMI cut-off for Asian Indians and defined lean or non-over-weight state as BMI < 23 kg/m2 [17, 18]. We felt the cut-off of BMI < 23 kg/m2, would more accurately define the lean body habitus than choosing the prevalent international cut-off for non-overweight state of BMI < 25 kg/m2.
Individuals with lean NAFLD, non-lean NAFLD and no NAFLD (none of the three USS criteria for NAFLD and no or ‘safe’ alcohol consumption) in 2007 were compared for baseline differences in gender, anthropometry [weight and waist circumference (WC)] and the presence of metabolic comorbidities such as diabetes, hypertension, hypertriglyceridemia and low-HDL levels. The three groups were also compared for the development of incident diabetes, hypertension, hypertriglyceridemia, low-HDL and body composition after 7 years of follow-up in 2014.
We also looked at risk factors for the development of new onset lean NAFLD in 2014. For this the sample was restricted to those who were lean at both baseline and follow-up, in 2007 and 2014, to ensure that the confounding effect on weight changes on the outcome over the 7 years of follow-up was limited as much as possible. Therefore, new onset lean NAFLD was defined as those developed new NAFLD in 2014 but who remained lean (BMI < 23 kg/m2) at baseline as well as in 2014 at follow-up, with ‘safe’ alcohol intake throughout. Non-NAFLD lean controls were those who had none of the USS criteria for NAFLD and had a BMI < 23 kg/m2 in both 2007 and 2014 with no or ‘safe’ alcohol intake throughout.
To test associations of selected genetic variants with lean NAFLD, we performed a case–control study between lean NAFLD case and lean non-NAFLD control subjects after the 7-year follow-up. Of 10 selected SNPs, 5 SNPs had been previously identified to be associated with NAFLD and related-phenotypes at a genome-wide significance level (p ≤ 5×10−8) in European genome-wide association studies [PNPLA3 (rs738409), LYPLAL1 (rs12137855), GCKR (rs780094), PPP1R3B (rs4240624) and NCAN (rs2228603)], and the remaining 5 SNPs were derived from 3 candidate genes previously documented for NAFLD [APOC3 (rs2854117, rs2854116), ADIPOR2 (rs767870) and STAT3 (rs6503695, rs9891119)] [13]. The genotype distribution of all tested SNPs was in Hardy–Weinberg equilibrium (at p > 10−3). Genetic association was tested between lean NAFLD cases and lean non-NAFLD controls among lean individuals (BMI < 23 kg/m2) at each timing of survey; 84 cases vs. 461 controls at baseline (N = 545), where the individuals were restricted to those who also attended follow-up, and 282 cases vs. 283 controls at follow-up (N = 565), where part of the individuals were not in the case–control sample at baseline. Genotyping and statistical analysis were performed as previously described [13].
Data were entered in Epi Info 7 (Centres for Disease Control and Prevention, Atlanta, GA, USA) and logical and random checks were done. Statistical analysis was done using Stata 14.1 (StataCorp, College Station, Texas, USA). Continuous and categorical data were described using mean and standard deviations and percentages, respectively. Bivariate analysis was done using the Chi-squared test. Multivariate analysis was done using binary logistic regression. p < 0.05 was considered as significant.
Results
There were 3012 participants in the initial study of whom 2985 (99.1%) had complete data for analysis (1636 female (54.8%), mean age (SD) 54.2 (7.8) years). Of the inception cohort, 2148 (72%) [1237 (57.6%) female, mean age (SD) 54.2 (7.7 years)] participated in the follow-up assessment in 2014. Except for fewer males attending follow-up, the other characteristics (mean BMI, percentage with diabetes, percentage with hypertension, mean TG and mean HDL) were similar among the inception and follow-up cohorts.
Of the 2985 participants in 2007, 120 (4.0%) had lean NAFLD and 816 (27.3%) had non-lean NAFLD. 1206 (40.4%) had no evidence of NAFLD (non-NAFLD) (Fig. 1). Table 1 compares the baseline characteristics of the lean NAFLD, non-lean NAFLD and no NAFLD groups. Compared to non-lean NAFLD, lean NAFLD was significantly more common among males, and had a lower prevalence of hypertension and central obesity at baseline. There was no significant difference in prevalence of other metabolic comorbidities between lean and non-lean NAFLD at baseline.
Of those who attended follow-up in 2014, 1362/2148 (63.4%) had NAFLD. Of these, 13.2% (284/2148) were lean NAFLD and 50.2% (1078/2148) were non-lean NAFLD. In 2014, the 2148 who presented for follow-up included 84 of those who had lean NAFLD in 2007, 620 who had non-lean NAFLD in 2007 and 834 who did not have features of NAFLD in 2007 (Fig. 1). Table 2 shows incident metabolic comorbidities of the three groups at follow-up after 7 years (in 2014). There were no significant differences in the occurrence of new onset metabolic comorbidities such as diabetes, hypertension and dyslipidemia, between lean and non-lean NAFLD. However, incident diabetes and low HDL were significantly higher in non-lean NAFLD compared to no NAFLD.
Of those who did not have NAFLD in 2007, 294 were lean in both 2007 and 2014, and had no or ‘safe’ alcohol consumption in the period of follow-up. 84 of them [male 36 (42.9%); mean age 59.4 (SD = 8.0) years] had developed NAFLD in 2014, giving an annual incidence of lean NAFLD of 4.1% (Fig. 1). The 210 lean individuals who did not develop NAFLD were considered as controls for regression analysis. On logistic regression, the presence of diabetes at baseline, an increase in weight from baseline to follow-up and an education above secondary level were independent risk factors for the development of new onset lean NAFLD (Table 3).
Genetic association with NAFLD was replicated for PNPLA3 rs738409 among lean individuals (one-tailed p < 0.05, Table 4). The strength of association with lean NAFLD appeared to be attenuated at follow-up compared to baseline (odds ratio = 1.37 and 1.51 at follow-up and baseline, respectively), which is in accordance with our previous findings in the whole cohort sample [13]. Moreover, PNPLA3 association was pronounced among lean individuals compared to the whole cohort sample [13]; OR = 1.51 (baseline) and 1.37 (follow-up) for lean individuals vs. OR = 1.48 (baseline) and 1.22 (follow-up) for the whole cohort sample. Although nominal (p < 0.05) association was observed for LYPLAL1 rs12137855 (OR = 1.45, p = 0.022) and NCAN rs2228603 (OR = 0.57, p = 0.049) among 10 selected polymorphisms, careful interpretation is required without replication in an independent sample, considering the risk of multiple testing.
Discussion
In a cohort of aging adults living in an urban Sri Lankan community, the prevalence of lean NAFLD was 4% in 2007 and 13.2% in 2014. After 7 years of follow-up the annual incidence of lean NAFLD was 4.1%. Lean NAFLD was commoner among males, and had a lower prevalence of hypertension and central obesity than their non-lean NAFLD counterparts at baseline. Lean NAFLD had a similar risk to non-lean NAFLD for development of incident metabolic comorbidities. The presence of diabetes at baseline, an education above secondary level and an increase in weight from baseline were risk factors for the development of new onset lean NAFLD. To our knowledge, this is the first prospective, community-based, cohort follow-up study that describes characteristics of lean NAFLD in a South-Asian population.
Previous studies on lean NAFLD have reported varying, and sometimes contradictory findings. Table 5 summarizes the studies on lean NAFLD in Western and Asian populations including the present study [6, 9, 10, 19,20,21,22,23,24,25,26,27,28]. Our reported prevalence of 13.2% (in 2014) for lean NAFLD seems to be compatible with previous studies. It is lower than the prevalence of 18.8% reported by Younossi et al. [6] and 15.2% reported by Nishioji et al. [28] but higher than 9.6% reported by Kim et al. [20] employing BMI < 25 kg/m2 cut-off to define lean NAFLD. We found a lower prevalence of hypertension and central obesity among lean NAFLD than their non-lean NAFLD counterparts and this is comparable to the lower prevalence of metabolic syndrome (MetS) and components of MetS in lean NAFLD compared to overweight-NAFLD reported by Cruz et al. [9].
In the present study, there were no significant differences in the occurrence of incident metabolic comorbidities such as new-onset diabetes, hypertension and dyslipidemia, between lean and non-lean NAFLD after a 7-year follow-up. This highlights the need for stringent follow-up of all NAFLD patients for the development of new metabolic comorbidities, even in the absence of general obesity, and despite the lean habitus being a minority among NAFLD patients. This finding of the risk of future development of metabolic comorbidities among individuals with lean NAFLD has also been reported previously. Nishioji et al. described metabolic factors to be associated with NAFLD, even in non-obese Japanese [28]. Feng et al. reported lean NAFLD being more strongly associated with diabetes, hypertension and metabolic syndrome than overweight-obese-NAFLD in a Chinese population [27]. However, there have been previous reports of lean NAFLD being less strongly associated with MetS and its individual components. Younossi et al. reported that lean NAFLD is associated with a decreased likelihood of having insulin resistance and hypercholesterolemia [6]. Cruz et al. also reported lean NAFLD having less MetS, individual components of MetS and IR compared to overweight-NAFLD [9].
In our population, hypertension, general obesity and central obesity were less in the prevalent lean NAFLD group. This may suggest a stronger genetic predisposition for lean NAFLD than for non-lean NAFLD. However, diabetes predicted development of future lean NAFLD, and suggests an association between lean NAFLD and metabolic syndrome, and the need to develop NAFLD screening protocols for diabetic patients even in the absence of general obesity. We found that higher educational level was independently predictive of development of lean NAFLD. In previous published large population-based studies, a higher educational level was associated with lower BMI or leaner body habitus and most educated individuals displayed lower rates of obesity [29, 30]. The reason for the higher educated individuals with lean habitus to develop NAFLD compared to their lower educated lean counterparts in our study remains largely unexplained, but may be due to genetic predisposition.
PNPLA3 rs738409 showed a tendency of association with NAFLD among lean individuals (one-tailed p < 0.05); interestingly, the strength of association was pronounced among lean individuals compared to the whole cohort sample, indicating the possibility of stronger genetic predisposition for NAFLD among lean individuals compared to non-lean (or obese) individuals. This is in accordance with the observation that the strength of association with lean NAFLD was attenuated at follow-up compared to baseline. A similar population-based study from Hong Kong demonstrated that a greater proportion of patients with non-obese NAFLD carried the variant PNPLA3 allele than those with obese NAFLD (78.4% vs. 59.8%) and the PNPLA3 polymorphism remained an independent risk factor for non-obese NAFLD [31].
The main strengths of our study are that it is a prospective, community-based, cohort follow-up study with a relatively large baseline population presenting for re-evaluation. Furthermore, NAFLD was diagnosed on stringent ultrasound criteria, and not on surrogate markers such as hepatic transaminases. It is the only Asian study to report the incidence of lean NAFLD in an at risk lean population. There are, however, several limitations. Inter-observer reliability between sonographers was not formally assessed before the study commenced, and information on alcohol consumption was based on self-reporting; this may have led to under-reporting with consequent overestimation of the prevalence and incidence of NAFLD. We could not account for the possible changes and fluctuations in the BMI value of the lean individuals with NAFLD (lean NAFLD) in between the period from baseline and follow-up.
In conclusion, although lean NAFLD constitutes a small proportion of NAFLD, its impact on public health is not negligible given the high prevalence of cryptogenic CLD in Sri Lanka. The risk of developing incident metabolic comorbidities seems to be similar to non-lean NAFLD. Individuals with lean NAFLD, therefore, also warrant careful evaluation and follow-up.
References
Chalasani N, Younossi Z, Lavine JE, et al. The diagnosis and management of non-alcoholic fatty liver disease: practice guideline by the American Association for the Study of liver diseases, American College of Gastroenterology, and the American Gastroenterological Association. Hepatology 2012;55(6):2005–2023
Sayiner M, Koenig A, Henry L, Younossi ZM. Epidemiology of nonalcoholic fatty liver disease and nonalcoholic Steatohepatitis in the United States and the rest of the world. Clin Liver Dis 2016;20(2):205–214
Amarapurkar DN, Hashimoto E, Lesmana LA, et al. Asia-Pacific working party for NAFLD: how common is non-alcoholic fatty liver disease in the Asia-Pacific region and what are the local differences? J Gastroenterol Hepatol 2007;22:788–93
Loomis AK, Kabadi S, Preiss D, et al. Body Mass Index and risk of nonalcoholic fatty liver disease: two electronic health record prospective studies. J Clin Endocrinol Metab 2015;101(3):945–952
Wattacheril J, Sanyal AJ. Lean NAFLD: an underrecognized outlier. Curr Hepatol Rep 2016;15(2):134–139 (Epub 2016 Apr 14)
Younossi ZM, Stepanova M, Negro F, Hallaji S, Younossi Y, Lam B, Srishord M. Non-alcoholic fatty liver disease in lean individuals in the United States. Medicine (Baltimore) 2012;91(6):319–327
Liu CJ. Prevalence and risk factors for non-alcoholic fatty liver disease in Asian people who are not obese. J Gastroenterol Hepatol 2012;27:1555–1560.
Seto WK, Yuen MF. Non-alcoholic fatty liver disease in Asia: emerging perspectives. J Gastroenterol 2017;52:164. https://doi.org/10.1007/s00535-016-1264-3
Cruz ACD, Bugianesi E, George J, Day CP, Liaquat H, Charatcharoenwitthaya P, Mills PR, Dam-Larsen S, Bjornsson ES, Haflidadottir S, Adams LA, Bendtsen F, Angulo P. 379 Characteristics and long-term prognosis of lean patients with nonalcoholic fatty liver disease. Gastroenterology 2014;146(5):909
Feldman A, Eder SK, Felder TK, Kedenko L, Paulweber B, Stadlmayr A, Huber-Schönauer U, Niederseer D, Stickel F, Auer S, Haschke-Becher E, Patsch W, Datz C, Aigner E. Clinical and metabolic characterization of lean caucasian subjects with non-alcoholic fatty liver. Am J Gastroenterol 2017;112(1):102–110
Singh SP, Kar SK. The lean NASH conundrum. J Diabetes Metab 2016;7:642 https://doi.org/10.4172/2155-6156.1000642
Dassanayake AS, Kasthuriratne A, Rindrajith S, et al. Prevalence and risk factors for non-alcoholic fatty liver disease among adults in an urban Sri Lankan population. J Gastroenterol Hepatol 2009;24:1284–1288
Niriella MA, Pathmeswaran A, De Silva ST, Kasturiratna A, Perera KR, Subasinghe SKCE, Kodisinghe SK, Piyaratna TACL, Vithiya K, Dassanayaka AS, De Silva AP, Wickramasinghe AR, Takeuchi F, Kato N, de Silva HJ. Incidence and risk factors for non-alcoholic fatty liver disease among adults: a 7-year follow-up study in an urban Sri Lankan community. Liver Int 2017;37(11):1715–1722
Kasturiratne A, Akiyama K, Niriella MA, Takeuchi F, Isono M, Dassanayake AS, de Silva AP, Wickremasinghe RA, Kato N, de Silva HJ. Association of genetic variants with non-alcoholic fatty liver disease in an urban Sri Lankan community. Liver Int 2015;35:676–979
Senanayake SM, Niriella MA, Weerasinghe SK, Kasturiratne A, de Alwis JP, de Silva AP, Dassanayake AS, de Silva HJ. Survival of patients with alcoholic and cryptogenic cirrhosis without liver transplantation: a single center retrospective study. BMC Res Notes 2012;5:663
Dehghan M, Merchant AT. Is biometrical impedance accurate for use in large epidemiological studies? Nutr J 2007;7:26
Fan JG, Saibara T, Chitturi S, Sung JJY, Kim BI. Chutaputti A: Asia-Pacific working party for NAFLD. What are the risk factors and settings for non-alcoholic fatty liver disease in Asia–Pacific? J Gastroenterol Hepatol 2007;22:794–800
Bhardwaj S, Misra A. Obesity, diabetes and the Asian phenotype. World Rev Nutr Diet 2014. https://doi.org/10.1159/000362308
Margariti E, Deutsch M, Manolakopoulos S, Papatheodoridis GV. Non-alcoholic fatty liver disease may develop in individuals with normal body mass index. Annal Gastroenterol 2012;25(1):45–51
Kim HJ, Kim HJ, Lee KE, Kim DJ, Kim SK, Ahn CW, Lim S, Kim KR, Lee HC, Huh KB, Cha BS. Metabolic significance of nonalcoholic fatty liver disease in nonobese, nondiabetic adults. Arch Intern Med 2004;164(19):2169–2175. https://doi.org/10.1001/archinte.164.19.2169
Singh SP, Nayak S, Swain M, Rout N, Mallik RN, et al. Prevalence of nonalcoholic fatty liver disease in coastal eastern India: a preliminary ultrasonographic survey. Trop Gastroenterol 2004;25:76–79
Chen CH, Huang MH, Yang JC, Nien CK, Yang CC, et al. Prevalence and risk factors of nonalcoholic fatty liver disease in an adult population of taiwan: metabolic significance of nonalcoholic fatty liver disease in nonobese adults. J Clin Gastroenterol 2006;40:745–752
Das K, Das K, Mukherjee PS, Ghosh A, Ghosh S, et al. Nonobese population in a developing country has a high prevalence of nonalcoholic fatty liver and significant liver disease. Hepatology 2010;51:1593–1602
Choudhary NS, Duseja A, Kalra N, Das A, Dhiman RK, et al. Correlation of adipose tissue with liver histology in Asian Indian patients with nonalcoholic fatty liver disease (NAFLD). Ann Hepatol 2012;11:478–486
Bhat G, Baba CS, Pandey A, Kumari N, Choudhuri G. Insulin resistance and metabolic syndrome in nonobese Indian patients with non-alcoholic fatty liver disease. Trop Gastroenterol 2013;34:18–24
Singh SP, Kar SK, Panigrahi MK, Misra B, Pattnaik K, et al. Profile of patients with incidentally detected nonalcoholic fatty liver disease (IDNAFLD) in coastal eastern India. Trop Gastroenterol 2013;34:144–152
Feng R-N, Du S-S, Wang C, et al. Lean-non-alcoholic fatty liver disease increases risk for metabolic disorders in a normal weight Chinese population. World J Gastroenterol 2014;20(47):17932–17940. https://doi.org/10.3748/wjg.v20.i47.17932
Nishioji K, Sumida Y, Kamaguchi M, et al. Prevalence of and risk factors for non-alcoholic fatty liver disease in a non-obese Japanese population, 2011–2012. J Gastroenterol 2015;50:95. https://doi.org/10.1007/s00535-014-0948-9
Hermann S, Rohrmann S, Linseisen J, et al. The association of education with body mass index and waist circumference in the EPIC-PANACEA study. BMC Public Health 2011;17(11):169. https://doi.org/10.1186/1471-2458-11-169
Devaux M, Sassi F, Church J, et al. Exploring the relationship between education and obesity. OECD J 2011;1:121–159
Wei JL, Leung JC, Loong TC, et al. Prevalence and severity of nonalcoholic fatty liver disease in non-obese patients: a population study using proton-magnetic resonance spectroscopy. Am J Gastroenterol 2015;110:1306–1314
Acknowledgements
This work was supported by grants from the National Center for Global Health and Medicine, Tokyo, Japan and the Ministry of Higher Education of Sri Lanka. We thank those who have continuously supported the Ragama Health Study. We also thank many physicians for their assistance in collecting the DNA samples and accompanying clinical information, extracting DNA and preparing the analytical data.
Author information
Authors and Affiliations
Contributions
MAN, NK, AK and HJdeS conceptualized and designed the study. AK, SdeS, MAN, ASD, APDeS, ARW and HJdeS were involved in the establishment of the Ragama Health Study cohort and its follow-up. KRP, SKCES, SKK, TACLP and KV collected data. TF and NK performed genotyping. TF, NK, AK analyzed the data assisted by MAN, AP and HJdeS. MAN, AK, NK and HJdeS prepared the manuscript. APDeS, ASD and ARW were substantially involved in revision of the manuscript. All authors checked the final manuscript before submission.
Corresponding author
Ethics declarations
Conflict of interest
Madunil Anuk Niriella, A. Kasturiratne, A. Pathmeswaran, S. T. De Silva, K. R. Perera, S. K. C. E. Subasinghe, S. K. Kodisinghe, T. A. C. L. Piyaratna, K. Vithiya, A. S. Dassanayaka, A. P. De Silva, A. R. Wickramasinghe, F. Takeuchi, N. Kato and H. J. de Silva declare no conflict of interest.
Ethical approval
Ethical approval for the study was obtained from the Ethical Review Committees of the Faculty of Medicine, University of Kelaniya and NCGM.
Informed consent
Informed consent was obtained from all participants in RHS.
Rights and permissions
About this article
Cite this article
Niriella, M.A., Kasturiratne, A., Pathmeswaran, A. et al. Lean non-alcoholic fatty liver disease (lean NAFLD): characteristics, metabolic outcomes and risk factors from a 7-year prospective, community cohort study from Sri Lanka. Hepatol Int 13, 314–322 (2019). https://doi.org/10.1007/s12072-018-9916-4
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12072-018-9916-4